We study the molecular mechanisms involved in assembly and function of translation initiation complexes involved in protein synthesis, using yeast as a model system to exploit its powerful combination of genetics and biochemistry for dissecting complex cellular processes in vivo. The translation initiation pathway produces an 80S ribosome bound to mRNA with methionyl initiator tRNA (tRNAi) base-paired to the AUG start codon. The tRNAi is recruited to the 40S subunit in a ternary complex (TC) with GTP-bound eIF2 to produce the 43S preinitiation complex (PIC) in a reaction stimulated by eIFs 1, 1A, 3 and 5. The 43S PIC attaches to the 5' end of mRNA, facilitated by cap-binding complex eIF4F (comprised of eIF4E, eIF4G, and RNA helicase eIF4A) and PABP bound to the poly(A) tail, and scans the 5 untranslated region (UTR) for the AUG start codon. Scanning is promoted by eIFs 1 and 1A, which induce an open conformation of the 40S and TC binding in a conformation suitable for scanning successive triplets entering the ribosomal P site (P-out) , and by eIF4F and other RNA helicases, such as Ded1, that remove secondary structure in the 5' UTR. AUG recognition leads to tighter binding of TC in the P-in state and evokes irreversible hydrolysis of the GTP bound to eIF2, dependent on the GTPase activating protein (GAP) eIF5, releasing eIF2-GDP from the PIC leaving tRNAi in the P site. After joining of the 60S subunit, producing the 80S initiation complex, the eIF2-GDP is recycled to eIF2-GTP by guanine nucleotide exchange factor (GEF) eIF2B for the next round of translation initiation. Biochemical evidence for conformational changes in the P site and mRNA entry channel evoked by AUG recognition in reconstituted PICs. It is thought that the scanning PIC assumes an open conformation and that AUG recognition evokes a closed state that arrests scanning with more stable Met-tRNAi binding (P-in state), with attendant displacement of the eIF1A-CTT from the P site and dissociation of eIF1 from the 40S subunit. We obtained physical evidence for these conformational rearrangements by comparing patterns of directed hydroxyl radical cleavage (DHRC) of rRNA by Fe(II)-BABE tethered to unique cysteine residues engineered in eIF1A, in PICs reconstituted with mRNA with AUG or near-cognate (AUC) start codons. Key rRNA residues in the P site displayed reduced cleavage in AUG versus AUC PICs, suggesting that accessibility of these rRNA residues is reduced by their increased interaction with Met-tRNAi in the P-in state. The cleavage data also indicate that AUG recognition evokes dissociation of eIF1 from its 40S binding site, a constricted conformation of the entry channel, and displacement of the eIF1A-CTT from near the P-site. Structures of yeast preinitiation complexes reveal conformational changes from mRNA scanning to start-codon recognition. In collobaration with Venki Ramakrishnan's and Jon Lorsch's groups, cryo-EM reconstructions of yeast PICs have been obtained representing different stages of the initiation pathway, at 3.5-6.1 angstrom resolution. These include 40S-eIF1-eIF1A complexes and partial yeast 48S PICs in open and closed conformations (py48S-open and py48-closed). These structures provide a wealth of new information about conformational changes occurring in the transition from scanning to AUG selection. Comparing the 40S-eIF1-eIF1A complex to the free 40S reveals rotation of the 40S head that might promote TC binding to form the 43S PIC. The py48S-open, formed using mRNA with AUC start codon, reveals an upward shift of the 40S head that widens the mRNA entry channel and opens its latch--consistent with our eIF1A/Fe(II)-BABE cleavage data--which should facilitate mRNA insertion into the binding cleft to form the scanning PIC. Moreover, the P site is widened and lacks tRNAi contacts with the 40S body present in canonical 80S-tRNAi complexes. By contrast, py48S-closed, formed with mRNA(AUG), reveals downward head movement that closes the latch, clamps the mRNA into the binding cleft, and fully encloses tRNAi in the P site. The eIF1A NTT assumes a structured conformation and interacts with the AUG:anticodon duplex, consistent with its role in stabilizing P-in. eIF1 is repositioned on the 40S and deformed to prevent a clash with tRNAi, likely as a prelude to eIF1 dissociation from the 40S subunit. Both py48S-open and -closed complexes reveal eIF2beta and portions of the eIF3 complex. The eIF3 trimeric subcomplex eIF3b-CTD/eIF3i/eIF3g-NTD resides on the subunit-interface surface of the 40S, and appears to lock mRNA into the 40S binding cleft. eIF2beta interacts with tRNAi and segments of eIF1 and eIF1A exclusively in py48S-open, which could stabilize binding of TC and eIF1 to the scanning PIC prior to AUG recognition. A portion of eIF2alpha domain-I projects into the entry channel and contacts the beta-hairpin of uS7/Rps5 and nucleotides just upstream of the AUG codon, including the key -3 nucleotide of the Kozak consensus sequence for efficient AUG selection, consistent with a proposed role for eIF2alpha in start codon recognition. Exit-channel beta-hairpin of 40S ribosomal protein Rps5/u7 is a critical determinant of efficient and accurate translation initiation. In py48S complexes described above, the beta-hairpin of 40S protein Rps5/uS7 protrudes into the mRNA exit-channel, contacting the TC and mRNA context nucleotides; however, its importance in AUG selection was unknown. We identified substitutions in beta-strand 1 (E144R) and a nearby C-terminal residue (R225K) of yeast Rps5 that reduced bulk initiation, conferred leaky scanning of an upstream AUG, and lowered initiation fidelity by exacerbating the effect of poor context of the eIF1 start codon to thereby reduce eIF1 abundance. After reinstating WT eIF1 abundance with an extra copy of its gene (SUI1), both Rps5 substitutions were found to suppress UUG initiation, consistent with their discrimination against the poor-context SUI1 AUG codon. Consistent with this, in vitro analysis of mutant ribosomes showed that E144R greatly destabilized P-in, increasing the dissociation rate (koff) of TC from reconstituted 43S-mRNA PICs with AUG or UUG start codons. Other substitutions in beta-hairpin loop residues also increased initiation fidelity, suppressing UUG initiation in cells expressing Sui- variant eIF2beta-S264Y; and one such mutant (R148E) destabilized P-in at UUG, but not AUG, start codons in PICs reconstituted with the eIF2beta-S264Y form of eIF2. Thus, the Rps5 beta-hairpin is crucial for efficient and accurate start codon recognition in vivo. Genome-wide analysis of translational efficiency reveals distinct but overlapping functions of yeast DEAD-box RNA helicases Ded1 and eIF4A. RNA helicases eIF4A and Ded1 are believed to resolve mRNA secondary structures that impede ribosome attachment to the mRNA or subsequent scanning to the start codon, but whether they perform unique or overlapping functions in vivo is poorly understood. We compared the effects of mutations in Ded1 or eIF4A on global translational efficiencies (TEs) in yeast by ribosome footprint profiling. Despite similar reductions in bulk translation, inactivation of a cold-sensitive Ded1 mutant substantially reduced the TEs of >600 mRNAs, whereas inactivation of a temperature-sensitive eIF4A mutant yielded <40 similarly impaired mRNAs. Ded1-dependent mRNAs exhibit greater than average 5UTR length and propensity for secondary structure, implicating Ded1 in scanning though structured 5' UTRs. Reporter assays confirmed that cap-distal stem-loop insertions increase dependence on Ded1 but not eIF4A for efficient translation. Our findings suggest that Ded1 is critically required to promote scanning through secondary structures in 5UTRs,while eIF4A promotes a step of initiation common to virtually all yeast mRNAs.